Figure 16

Any adverse events (treatment-related): Type I-II diabetes.

Four trials compared sildenafil to placebo in participants with depression.^{79, 91, 115, 167} Of these, three trials^{91, 115, 167} involved participants with major depressive disorder (MDD) in remission. In two of these trials, ^{91, 167} participants had ED associated with the use of antidepressants, namely, selective serotonin reuptake inhibitors (SSRIs). In the third trial, ¹¹⁵ about 28 percent of the participants used SSRIs. The three trials^{91, 115, 167} were deemed to be potentially combinable. Given that the fourth trial⁷⁹ looked at ED patients who had depression but were not being treated with antidepressants at the time of their enrollment (this being one of the exclusion criteria), SSRI use could not be considered the cause of the ED, and the trial was therefore not combined with the other three for meta-analysis.^{91, 115, 167}

The results for the following efficacy outcomes (i.e., numerical effect estimates and standard deviations [SDs]) were ascertained from the three trials: percentage of successful intercourse attempts, ^{115, 167} patients with improved erection (GEQ-Q1), ^{115, 167} and mean IIEF-Q3/Q4 score.^{91, 115} The mean IIEF-Q3/Q4 scores and the corresponding SDs were ascertained graphically from one trial.¹¹⁵ Separate meta-analyses for these efficacy outcomes are presented (see Figures 16–19). No meta-analysis for adverse events could be performed, due to a lack of sufficient detail for the adverse events definitions provided in the trials. Note that one trial⁹¹ included younger patients (mean: 45, range 18–55 years) compared with the other trial (mean: 53, range 24–75 years).¹¹⁵